Interesting, I have been saying that for years, didn't know there was an entire paper on it.
The catch is only that because vented systems are an higher order system (= more variables), they are far more prone to tolerances and inaccurate behavior.
Practically speaking they are also much harder to tune and non-linearities like BL(x) and Cms(x) will really throw everything completely off.
In a simple way this means that the practical results won't be the same as what is expected.
The catch is only that because vented systems are an higher order system (= more variables), they are far more prone to tolerances and inaccurate behavior.
Practically speaking they are also much harder to tune and non-linearities like BL(x) and Cms(x) will really throw everything completely off.
In a simple way this means that the practical results won't be the same as what is expected.
It's unpublished, except that I made it available informally on some audio forums.
I’ve wanted to try increasing compliance by burning off some of the spider with a soldering iron. I got too skeerd to try it.
Yep.
Drivers are predominantly minimum phase.
Speakers with xovers have frequency regions where they are minimum phase,..... along with regions where they are not minimum phase.
Plain and simple imo.
Hi @lrisbo,
Kindest regards,
M
Not implying that you are incorrect, but is there any paper that would describe the theory behind your claim? I am asking because I do have a low Qts (0.2) woofer, and when I inquired about the application, a huge enclosure was recommended by several people.
Kindest regards,
M
There is not one single system in this universe that is 100% minimum phase.
After a while ALL of them become less linear.
Some more than others.
That whole idea is called control-theory.
But this is nothing new either?
It's being dramatic about the obvious.
I haven't been referring to non-linear behavior as a source of minimum phase disruption..
As has been said....the source is crossovers....their regions of phase rotation.
Or going back to a favorite, group delay, LOL. A speaker is minimum phase only in the regions of flat group delay.
A speaker is not minimum phase where group delay is not flat.
Here's a simple two-way electrical summation of LR24's @ 1000Hz.
The region of the group delay's slope and curvature, between say about 200Hz and 5kHz, is NOT minimum phase.
That electrical region, knocks out minimum phase for the speaker for same range, if so crossed.
As far as I know, minimum phase means casual and stable (incl its inverse).
Which is as far as I know always applicable within the linear range.
Otherwise a system becomes non-linear.
But i have to admit that this is not only a matter of how these words are being used, but there are also a few translation issues here between languages.
A sealed box with a given Qtc has a corner frequency of fc=fs•Qtc/Qts. The needed box volume is something like Vas/(Qtc/Qts-1) so if Qts approaches the desired Qtc then the Volume goes to infinity. A further analysis shows that fc is independent of the suspension compliance. However, the high the compliance the smaller box for a given fc and Qtc. And a higher compliance gives a lower Qts. Hence drivers for acoustic suspension have low Qts
Linkwitz Transform is a parametric EQ in which the resonant frequencies of the denominator and the numerator are independently adjustable, in addition to the gains and the Qs.
Last edited:
What would it be like to leave out all causally non-linear components when developing a loudspeaker?
Thread title: "Why are sealed box woofers out of fashion" (?)
Because there MUST be the causally linear Helmholtz port ! (@b_force)
the zeros (numerator) of the LT is designed to match (ie cancel) the poles (denominator) of the speaker response. A new and adjustable set of poles are in the denominator of the LT. This can effectively shift the speaker corner frequency up or down and change the alignment to anything. This also works for reflex (4 poles/4th order)
we can build an all-pass filter using LCR components. Consequently, just because it can be modelled with lumped components does not guarantee its minimum phase.
Regarding Linkwitz transform: This Kind of filter can also be called a biquad. There are more possible circuit solutions to achieve its transfer function. One of them is the use of a state variable filter with its three outputs weighted and added. The fourth order version could be used for the correction of reflex boxes.
Regards
Charles
Regards
Charles
Yes I agree
As we discovered a few posts ago, it's a bit a matter of definition.
As long as it is casual, stable and predictable.
But you are referring to the phase component as well.
Which can be tricky, because how would you define an entire system if it's phase compensated?
Besides the fact that it's not a hard discrete step between minimum phase and not.
Anyway, seeing from a practical point of view, we are getting very pedantic at this point.
Which is totally fine with me, but don't know if other people find it very relevant?